W18 · May 2026

The week AI got serious about DAM

Three things converged this week, and they all point in the same direction. AI is finally hitting the asset library, and the libraries that were not built for it are going to feel it first.

topic/ai-readinessAdobe Experience Cloud · May 1, 2026

Adobe sets Q3 timeline for Firefly inside Experience Manager ↗

Adobe today announced the next phase of Firefly inside Experience Manager Assets, with metadata auto-tagging arriving in a Q3 release. The integration extends Firefly's generative capabilities directly into AEM's asset workflow, allowing creative teams to generate variants and apply structured metadata in a single pipeline.

Read on Adobe Experience Cloud →

TL;DR · Why it matters

The headline is image generation. The thing to watch is the auto-tagging. AEM customers have been waiting on this since 2024, and the value depends entirely on what the AI tags an asset with. If those tags do not match the words your team actually uses to find things, you end up with a tidier-looking mess. This is a metadata structure question more than a software question.

Permalink & share

topic/metadataIPTC · April 28, 2026

IPTC publishes updated field standard for AI training data ↗

The International Press Telecommunications Council (IPTC) has published an updated photo metadata standard adding a new structured field for AI training data provenance. The revision identifies digital asset management systems as the canonical source for descriptive, rights, and lineage metadata used in machine learning datasets.

Read on IPTC →

TL;DR · Why it matters

IPTC names DAM as the primary source for AI training data. That sounds bureaucratic, but it changes the buying conversation: every team using AI on their content now has a reason to care about how their library is structured. If the fields are loose, training quality drops with them. The basics of a working metadata model are harder to ignore than they were last year.

Permalink & share

topic/governanceCMSWire · April 30, 2026

Three enterprise teams name approval workflows the biggest AI bottleneck ↗

A survey of three Fortune 500 enterprise digital asset programs identified asset approval workflows as the single largest bottleneck in their AI content pipelines. Across all three teams, average time-to-approval grew over the past 18 months even as upstream content generation accelerated.

Read on CMSWire →

TL;DR · Why it matters

Approval bottlenecks are not a software problem. They are a design problem. The teams that solve this do not add more reviewers, they remove the decisions that did not need a person looking at them in the first place. Mapping which decisions actually need a human review is the part most programs skip.

Permalink & share

Resource of the week

How to tell if your DAM is actually working

A six-question diagnostic for operators. Sixty seconds start to finish, with a longer guide if you want the deep read.

Run the diagnostic →

Field Notes

JobBynder is hiring a Director of Customer Success in Amsterdam.
ConferenceHenry Stewart DAM NY opens its CFP this week.
VendorCanto rolled a quiet update to its rights-management module.
ToolAir shipped a public API for taxonomy export.

House ad

Pull up a seat at DAM Cowboy.

Weekly intelligence, working tools, and vetted directories. Free.

Browse the field →

Reply and tell us what you are working on. We read every one.